EGF and NRG induce phosphorylation of HER3/ERBB3 by EGFR ... · recruitment of HER3 into EGF-dependent EGFR clusters re-sults in its phosphorylation through lateral propagation and

EGF and NRG induce phosphorylation of HER3/ERBB3by EGFR using distinct oligomeric mechanismsBettina van Lengericha, Christopher Agnewa, Elias M. Puchnerb, Bo Huangc,d, and Natalia Juraa,e,1

aCardiovascular Research Institute, University of California, San Francisco, CA 94158; bSchool of Physics and Astronomy, University of Minnesota, Twin Cities,Minneapolis, MN 55455; cDepartment of Pharmaceutical Chemistry, University of California, San Francisco, CA 94158; dDepartment of Biochemistry andBiophysics, University of California, San Francisco, CA 94158; and eDepartment of Cellular and Molecular Pharmacology, University of California, SanFrancisco, CA 94158

Edited by Susan S. Taylor, University of California at San Diego, La Jolla, CA, and approved February 23, 2017 (received for review November 1, 2016)

Heteromeric interactions between the catalytically impaired hu-man epidermal growth factor receptor (HER3/ERBB3) and itscatalytically active homologs EGFR and HER2 are essential fortheir signaling. Different ligands can activate these receptor pairsbut lead to divergent signaling outcomes through mechanismsthat remain largely unknown. We used stochastic optical recon-struction microscopy (STORM) with pair-correlation analysis toshow that EGF and neuregulin (NRG) can induce different extentsof HER3 clustering that are dependent on the nature of thecoexpressed HER receptor. We found that the presence of theseclusters correlated with distinct patterns and mechanisms ofreceptor phosphorylation. NRG induction of HER3 phosphorylationdepended on the formation of the asymmetric kinase dimer withEGFR in the absence of detectable higher-order oligomers. UponEGF stimulation, HER3 paralleled previously observed EGFR behav-ior and formed large clusters within which HER3 was phosphory-lated via a noncanonical mechanism. HER3 phosphorylation byHER2 in the presence of NRG proceeded through still anothermechanism and involved the formation of clusters within whichreceptor phosphorylation depended on asymmetric kinase dimer-ization. Our results demonstrate that the higher-order organizationof HER receptors is an essential feature of their ligand-inducedbehavior and plays an essential role in lateral cross-activation of thereceptors. We also show that HER receptor ligands exert uniqueeffects on signaling by modulating this behavior.

HER/ERBB receptors | receptor tyrosine kinase signaling | receptorclustering | STORM | EGFR activation

The human epidermal growth factor receptors (HERs/ErbBs)are essential regulators of development and adult homeo-

stasis (1). All four of them, EGF receptor (EGFR; HER1),HER2 (ErbB2), HER3 (ErbB3), and HER4 (ErbB4), are at thefocus of therapeutic efforts in a variety of human diseases. Mostof what we know about their activation mechanism has beenrevealed by the studies on EGFR, which showed that ligandbinding induces EGFR dimerization through a series of struc-turally well-defined interactions between the extracellular andintracellular receptor domains (2). These interactions result inthe formation of an asymmetric kinase dimer in which one kinase(termed the “activator kinase”) is asymmetrically positioned toactivate the second kinase (termed the “receiver kinase”) allo-sterically (3). The receiver kinase then is poised to phosphorylatethe receptor tails, resulting in the recruitment of downstreamsignaling molecules and signal propagation.One of the characteristic features of the HER receptor family

is a significant degree of heteromeric interactions in response toligand binding through which the receptors activate a variety ofsignaling pathways (1). These interactions are particularly im-portant for signaling by the orphan receptor HER2 and thecatalytically impaired HER3, which do not signal on their ownunder normal conditions. Although all HER receptors are as-sumed to form heterodimers in which the kinase domains re-capitulate the asymmetric kinase homodimer characterized forEGFR (3), the molecular details of the heteromerization are

largely not understood. The protein interfaces involved in theasymmetric kinase domain interactions are highly conservedamong all HER receptors and have been structurally shown tosupport the formation of another active HER receptor homo-dimer, HER4/HER4, as well as an EGFR/HER3 heterodimer(4, 5). Enzymatic studies on the isolated kinase domains of HERreceptors have also provided convincing evidence that theircatalytic activation in heterodimers is dependent on the asym-metric dimer interface (4–7). On the other hand, because of alack of structures of liganded extracellular domain heterodimers,we do not know how ligand binding promotes heterotypic in-teractions between the extracellular portions of these recep-tors. Functional studies in cells have shown that binding of aligand cognate to only one HER receptor is sufficient to inducecross-activation of other HER receptors (8–10). How these in-teractions are further fine-tuned by ligands with different spec-ificity for the individual receptors is presently not completelyunderstood.Several studies have indicated that heterodimeric HER re-

ceptor complexes are functionally distinct, depending on whichcognate ligand activates them. The example of pairing betweenEGFR, which specifically binds its own ligands such as EGF, andthe catalytically impaired HER3, which binds its own ligandneuregulin (NRG), is particularly intriguing. Signaling crosstalkbetween EGFR and HER3 plays a predominant role in signalingin the adult liver and in melanomas (11, 12) and in underlyingpoor response to therapies that directly target HER2 (13). TheEGFR/HER3 pair can be activated either through EGF or

Significance

Signaling by human epidermal growth factor receptor (HER)receptors relies on heteromeric interactions between fourmembers of the family: EGF receptor, HER2, HER3, and HER4.These interactions remain remarkably poorly understood. Us-ing super-resolution microscopy imaging and signaling assays,we demonstrate a rich scope of HER receptor organizationpatterns that are differentially influenced by ligands and co-receptor expression, resulting in unique phosphorylation sig-natures of HER receptors. We also show that there arefundamental differences in molecular mechanisms that governHER receptor cross-activation, which do not always follow thecanonical kinase dimerization mechanism. Our data underscorean emerging concept in the field that HER receptor signalingneeds to be interpreted in the context of higher-order receptoroligomers, redefining the basic signaling unit relevant forreceptor function.

Author contributions: B.v.L., C.A., and N.J. designed research; B.v.L. and C.A. performedresearch; E.M.P. and B.H. contributed new reagents/analytic tools; B.v.L., C.A., E.M.P., B.H.,and N.J. analyzed data; and B.v.L., C.A., and N.J. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1To whom correspondence should be addressed. Email: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1617994114/-/DCSupplemental.

E2836–E2845 | PNAS | Published online March 20, 2017 www.pnas.org/cgi/doi/10.1073/pnas.1617994114

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020

NRG, but although the same active enzyme (EGFR) is engagedin both cases, the resulting pattern of receptor phosphorylation issignificantly different when the EGFR/HER3 complex is in-duced by EGF or NRG stimulation (Fig. 1). EGF induces robustphosphorylation of both receptors, whereas NRG stimulationresults in poor EGFR phosphorylation (Fig. 1) (10, 14). Basedon the structural understanding of the EGFR/HER3 kinase in-teraction, these two kinases should interact in the same way ineach case, with EGFR in the receiver position and HER3 servingas an allosteric activator. What then explains differences inphosphorylation outcomes?Although our mechanistic understanding of ligand-induced

HER receptor activation has been developed primarily in thecontext of a dimer model, initial studies on the behavior ofEGFR in response to EGF referred to higher-order oligomeri-zation (15, 16). Subsequently, numerous studies have providedquantitative evidence that ligand-induced active EGFR com-plexes extend beyond dimers to larger oligomers. In some stud-ies, such indications came from the experiments in which FRETwas detected between two EGF ligand molecules; because of thestructural constraints of their interaction, this transfer is mostlogically explained by receptor oligomerization (17–21). Otherstudies imaged fluorescently labeled EGFR itself using imagecorrelation microscopy (ICS) (22, 23), fluorescence correlationspectroscopy (FCS) (24, 25), stochastic optical reconstructionmicroscopy (STORM) (26), FRET, fluorescence lifetime imag-ing microscopy (FLIM) (27, 28), spatial mapping of the receptorby immuno-electron microscopy (29, 30), or number andbrightness analysis (31). This spectrum of experimental ap-proaches all led to a unifying conclusion that EGFR is organizedin larger clusters in response to ligand binding. More recentstudies provide evidence that these higher-order multimers areneeded to achieve the complete spectrum of EGFR phosphor-ylation (21, 25).At present, it is uncertain whether in the HER family higher-

order oligomerization is unique to EGFR or constitutes an es-sential component of signaling by all HER receptors. It is alsounknown how this behavior of EGFR influences its interactionswith other HER receptors and to what extent the formation ofhigher-order complexes is involved in HER receptor crosstalkinitiated by different ligands. In this study we set out to un-derstand the mechanism by which EGF and NRG stimulatedifferent phosphorylation states of HER receptors. UsingNR6 cells that do not express detectable levels of any HER re-ceptors, we created stable cell lines expressing HER receptors

and imaged these receptors in the range corresponding tophysiological expression levels. We applied STORM analysisthat combines conventional STORM superresolution imagingwith a blinking correction algorithm. In STORM, the position ofa single molecule can be determined with a precision of ∼20 nm(FWHM) over a wide range of receptor densities. As a result,this method provides an advantage over published single-molecule imaging analyses of HER receptors, which often arebased on single-molecule tracking of a sparsely labeled HERreceptor or its ligand and are confounded by contributions fromunlabeled “dark” receptors that represent the endogenous re-ceptor populations (17–22, 32, 33). Additionally, because manyfluorophores commonly used for STORM blink erratically (34),we chose a fluorophore [a photo-activatable fluorescent protein(PAFP), mEos3.2, hereafter referred to as “mEos”] that enablescorrection of the blinking behavior in postimaging analysis. Thecorrection eliminates the overcounting of receptors, which isinevitable when the quantification of receptor oligomerization isbased on fluorescence intensity, as applied in previous STORManalysis of EGFR oligomerization (26). We then scored theblink-corrected molecular positions with a pair-correlationanalysis, obtaining a reliable representation of the extent of re-ceptor clustering based on the fluorescent image.We show that when EGFR and HER3 are coexpressed, EGF

induces robust phosphorylation of both receptors, but NRGstimulation leads only to HER3 phosphorylation. Using STORMwith blink-correction analysis, we show that in response to thetwo different ligands, EGF and NRG, EGFR and HER3 popu-late distinct oligomeric states, with EGF stimulation resulting inhigher-order clustering of both receptors driven by EGFR olig-omerization. Moreover, we show that although phosphorylationof HER3 in response to NRG is a consequence of direct asym-metric dimer formation between EGFR and HER3, the asym-metric interaction between these two receptors is dispensable fortheir phosphorylation in response to EGF. Our data indicate thatrecruitment of HER3 into EGF-dependent EGFR clusters re-sults in its phosphorylation through lateral propagation and isindependent of the formation of the canonical active complexbetween the EGFR and HER3 kinase domains. In contrast,HER3 interaction with HER2 in response to NRG binding alsoproceeds through cluster formation but remains dependent onthe canonical asymmetric interactions within the HER2/HER3 kinase dimer. Our data thus unravel unique features ofmembrane organization of HER receptors and a surprisingcomplexity of the mechanisms that support ligand-inducedcrosstalk in the HER receptor family.

ResultsEGF and NRG Induce Different Phosphorylation Receptor States in theEGFR/HER3 Complexes. To examine the differences in the mecha-nism for receptor activation when signaling by the EGFR/HER3 complex is induced by EGF or by NRG at normal phys-iological levels of receptor expression, we created stableNR6 cell lines expressing HER receptor constructs. NR6 cells donot express detectable levels of HER receptors, minimizing theinterference from endogenous HER receptors (see Fig. 6 forHER2 and Fig. S1A for EGFR and HER3). Stable cell lineswere sorted by flow cytometry (Fig. S1B), and we subsequentlyused STORM to image selectively cells that express 20–60 re-ceptors per square micrometer on average, corresponding to theaverage density of the receptor observed under normal physio-logical conditions (Fig. S1C) (35). The receptors were taggedwith mEos to allow quantitative analysis of HER receptorsby STORM in later analyses. We fused mEos between theN-terminal signal peptide and the beginning of the mature ex-tracellular domain of HER receptors (Fig. 2A). The mEos fusiondid not affect the normal function of HER receptors in responseto ligand binding (Fig. S1D).To evaluate the extent of EGFR and HER3 activation upon

stimulation with EGF or neuregulin 1β (NRG1β; referred to as“NRG” throughout the paper), we measured the phosphorylation

pY1289 HER3

HER3

EGFR

pY1068 EGFR

pY1197 HER3

pY1222 HER3

pY1045 EGFR

pY1173 EGFR

pS473 AKT

AKT

ERK

pT202/Y204ERK

-EGF NRG

2 5 10 60 2 5 10 60

mEos-HER3:EGFR-GFP: + + + + + + + + +

+ + + + + + + + +

[min]:

-EGF NRG

2 5 10 60 2 5 10 60

mEos-HER3:EGFR-GFP: + + + + + + + + +

+ + + + + + + + +

[min]:

EGFR-GFP

mEos-HER3

EGF

NRGor

x

mEos

GFP

Fig. 1. Differential effects of NRG and EGF on the phosphorylation ofcoexpressed HER3 and EGFR. NR6 cells stably coexpressing mEos–HER3 andEGFR–GFP were serum starved for 6 h, stimulated with 10 nM EGF or 10 nMNRG for 2–60 min, and subsequently lysed. Western blot analysis of the ly-sates with the indicated antibodies reveals phosphorylation states of EGFRand HER3 as well as AKT and ERK. Antibodies used are listed in Table S1.

van Lengerich et al. PNAS | Published online March 20, 2017 | E2837

BIOCH

EMISTR

YPN

ASPL

US

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020

200

150

100

50

0

HE

R3

coun

ts

100806040200

distance (nm)

no EGF 1 min EGF 5 min EGF 10 min EGF

A B

r

combine blinks9.2

9.0

8.8

8.6

8.4

y-p

ositi

on (

m)

13.012.812.612.4

x-position ( m)

40x103

30

20

10

0

Fram

e number

C monomer oligomer

r

20nm

500

400

300

200

100

0

2

100806040200

distance (nm)

500

400

300

200

100

0

2

100806040200

distance (nm)

kinasedomain

C-terminaltail

extracellulardomain

mEos

signal peptide

Y

Y

Y

D

0 min 1 min 10 min 5 min

1 m

0 min 1 min 10 min 5 min

H

G

E

F I

200

150

100

50

0

HE

R3

coun

ts

100806040200

distance (nm)

no NRG 1 min NRG 5 min NRG 10 min NRG

200

150

100

50

0

EG

FR

cou

nts

16012080400

distance (nm)

no EGF 1 min EGF 5 min EGF 10 min EGF

200

150

100

50

0

HE

R3

coun

ts

100806040200

distance (nm)

no NRG 1 min NRG 5 min NRG 10 min NRG

mEos-EGFR

EGF

mEos-HER3

NRG

x

NRG

x

mEos-HER3mEos-HER3

x

EGF

mEos mEos mEos mEos

400

300

200

100

0

inte

grat

ed c

ount

s

none 1 min 5 min 10 min

EGFR+EGF HER3+NRG

***

***

***

400

300

200

100

0

inte

grat

ed c

ount

s

none 1 min 5 min 10 min

EGF NRG

***

**

***

EGFR-GFPEGFR-GFP

GFP GFP

mEOS-EGFR + EGF

mEOS-HER3 + NRG

mEOS-HER3/EGFR-GFP + EGF

mEOS-HER3/EGFR-GFP + NRG

distance (nm)distance (nm) distance (nm) distance (nm)

HE

R3

coun

ts

HE

R3

coun

ts

HE

R3

coun

ts

EG

FR

cou

nts

Fig. 2. STORM imaging of EGFR and HER3 upon ligand stimulation. (A) Schematic representation of the mEos–HER receptor fusion. (B) Fluorescent blinks(colored crosses) from individual mEos-tagged receptors were combined based on a threshold radius (r) and time (color scale). The positions of all blinks wereaveraged to determine the blink-corrected position (denoted by a black x). (C) Molecular positions of a simulated monomer or oligomer. The number ofmolecules within each red shell (which has a thickness of the bin size, 20 nm) was counted and normalized for the area of each shell for each molecule in theimage. Counts are summed in the pairwise distance histograms and show increasing height of bins at short distances with increasing size of oligomers becauseof the increase in the local density of molecules. Red curves reflect the bin height and are plotted instead of bins in all subsequent figures to allow simplerplotting of multiple overlapping histograms. (D, Upper) Cartoon schematic of the constructs and ligands used in the experiment. (Lower) Representativereconstructed STORM images of mEos–EGFR and mEos–HER3 organization at the plasma membrane of NR6 cells after stimulation with 10 nM EGF or 10 nMNRG for the indicated amount of time. Cells were imaged using TIRF illumination to detect receptors selectively at the plasma membrane. (E) STORM imagesof mEos–EGFR or mEos–HER3 under the indicated conditions were corrected for blinking as described in B; then the pairwise distance histograms wereconstructed from the blink-corrected molecular positions. For all experiments, the count values for each bin are the median of hundreds of nonoverlappingregions of interest (25 × 25 pixels) analyzed in total and were compiled from 10–12 cells; all error bars reflect the SE. Data were compiled from all experiments,including experiments performed on different days, to account for sample variability. Statistical significance for all plots was calculated using a two-sidedWilcoxon rank sum test for equal medians. (F) Sum of the counts (y-value, also termed “integrated counts” here) from the pairwise correlation histograms in Eup to 80 nm (first four bins), representing the increase in total number of receptors above the average density within a radius of 80 nm. Error bars representthe sum of the four SEs of the bins that were summed. ***P < 0.001. (G) Cartoon schematic (Upper) and STORM images (Lower) of mEos–HER3 coexpressedwith EGFR–GFP and stimulated with ligand (EGF or NRG) for the indicated times. (H) Pair-correlation analysis as described in E for STORM images of mEos–HER3 coexpressed with EGFR–GFP and stimulated with EGF or NRG. (I) Sums of counts calculated as in F for the above H pair-correlation histograms. **P <0.01; ***P < 0.001.

E2838 | www.pnas.org/cgi/doi/10.1073/pnas.1617994114 van Lengerich et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020

status of key tyrosine residues in the C-terminal tails of EGFR andHER3 by Western blot analysis of the NR6 cell lines stablycoexpressing tagged HER3 and EGFR receptors. As shown inFig. 1, HER3 phosphorylation is comparable when cells cotrans-fected with HER3 and EGFR are stimulated with EGF or NRG.This HER3 phosphorylation translates to similar levels of serine-threonine kinase (AKT) phosphorylation, which is predominantlyinduced by phosphorylated HER3 (Fig. 1). In contrast, althoughphosphorylated EGFR is robustly detected in response to EGF,almost no signal is detected in response to NRG (Fig. 1). Con-sequently, phosphorylation of ERK, a main downstream target ofEGFR, is also significantly lower in response to NRG than withEGF stimulation (Fig. 1).The observed lack of EGFR phosphorylation in response to

NRG is in agreement with previous observations (10, 14), but itis somewhat puzzling in the context of our understanding of themechanism of HER receptor activation. According to thismechanism, EGFR and HER3 are expected to form an asym-metric kinase dimer upon ligand-induced heterodimerization.Because of severe impairment in catalytic activity, HER3 hasbeen shown to act as the allosteric activator of EGFR when thesetwo partners form an asymmetric dimer (6, 7). Because EGFR isthe only active kinase in this complex, the resulting receptorphosphorylation is a result of its activity. Although phosphory-lation of EGFR in cis in the active dimer has been shown to beless pronounced than in trans, it does occur (36). Why thencannot EGFR autophosphorylate itself when stimulated withNRG in the presence of HER3?

STORM-Based Measurement of Receptor Clustering. To understandthe underlying mechanism for the difference in the response ofthe coexpressed EGFR and HER3 receptors to stimulation withtheir cognate ligands, we examined how ligand addition affectstheir spatial organization at the plasma membrane using ourpair-correlation analysis of blink-corrected STORM images. Theuse of mEos as a fluorescent label of HER receptors lies at thecore of our STORM analysis. Unlike organic fluorophores suchas Alexa 647, whose blinking behavior can be erratic and thussubject to overcounting, some PAFPs, such as mEos, exhibit abrief burst of blinks and then photobleach and do not return to abright state (see Fig. S2A for a representative fluorescence timetrace). The blinking behavior of mEos then can be deconvolutedby applying a blink-correction algorithm to discern whetherfluorescent bursts originated from a single molecule or frommultiple molecules within the spatial resolution (20 nm) (34). Inthis method, bursts of fluorescence (blinking) originating from asingle molecule can be combined by applying thresholds in boththe temporal and spatial dimension of the fluorescence locali-zations (Fig. 2B). Both thresholds were determined experimen-tally: the temporal threshold from the dark-state lifetimedistribution, and the spatial threshold from the pair-correlationhistogram of the raw data (Fig. S2 B and C). The pair-correlationhistogram is the distribution of distances between any two pairsof localizations; this function is flat if all molecules are distrib-uted randomly but exhibits a peak with increased local density ofmolecules around another molecule (Fig. 2C). The amplitude ofthe peak increases if the local density increases, and the widthof the peak increases with cluster size (Fig. 2C). After combiningthe blinks originating from single molecules, we constructed ablink-corrected pair-correlation histogram, which we used as aquantitative readout of receptor clustering.

When Expressed Alone, EGFR, but Not HER3, Clusters in Response toBinding of the Cognate Ligand. Using STORM, we first looked athow the organization of EGFR and HER3 at the plasmamembrane changes in response to stimulation with their re-spective selective ligands, EGF and NRG, when these receptorsare expressed alone. NR6 cells stably expressing mEos-taggedEGFR or HER3 were serum starved, stimulated with their re-spective ligands for varied periods of time, and subsequentlyfixed and imaged by STORM using total internal reflection

fluorescence (TIRF) illumination to image the receptors selec-tively at the plasma membrane. Upon stimulation with EGF, weobserved that EGFR molecules start to cluster as early as 1 minafter EGF stimulation, forming increasingly larger clusters atlater time points (Fig. 2 D–F). Using a counting radius of 50 nm(chosen because it is approximately equal to the FWHM spatialresolution of the raw STORM image), we created histograms ofthe number of receptors per cluster and estimated that theselarger clusters contain ∼5–20 receptor molecules on average(Fig. S3A). This number is an estimate, and the exact numbercan depend on several parameters and aspects of the countingmethod; therefore we use it only as a general comparison be-tween different receptor clusters. In contrast to EGFR, nomeasurable change in clustering could be detected for the cat-alytically impaired HER3 in the presence of NRG (Fig. 2 D–F).Notably, the pair-correlation histograms describing EGFR and

HER3 organization under the serum-starved conditions are notflat, suggesting that these receptors could exist in oligomericstates larger than monomeric in the basal condition (Fig. 2E). Tounderstand the origin of this basal-level peak in the pair-correlation, we used a calibration system previously applied tocount the number of molecules in clusters in yeast cells (34). Inthis method, different oligomeric states are modeled by fusion ofan increasing number of mEos molecules to the pleckstrin ho-mology (PH) domain of AKT, which is monomeric and localizesto the plasma membrane (34). We analyzed blink-corrected pair-correlation functions of 1×, 2×, and 3× repeats of mEos fusedto the PH domain in NR6 cells. In contrast to the analysis per-formed in yeast, the correlation function of the monomericPH-1×–mEos fusion in mammalian cells was not flat, even at rela-tively low (20 molecules/μm2) expression level (Fig. S2D).As an additional control, we also analyzed the membrane

organization of a single-pass transmembrane protein, CD86[previously demonstrated to be monomeric (37)], tagged at the Cterminus with mEos, and observed a similar peak in the pair-correlation function (Fig. S2E). Although there could be severalexplanations for this peak observed for both PH-1×–mEos andCD86–mEos, which did not appear in the counting study in yeast(34), it is possible that this behavior reflects intrinsic heteroge-neity of the mammalian cell membrane environment, such as thepresence of cholesterol-rich lipid microdomains or regions of themembrane that are not completely flat. These irregularitieswould be expected to prevent uniform distribution of membraneproteins. We therefore have refrained from interpreting thebasal state of HER receptor clustering at the plasma membraneand instead focused on relative changes in HER receptor orga-nization at the membrane in response to ligand stimulation.

EGF Induces HER3 Clustering in the Presence of EGFR. We next usedSTORM to investigate the effect of EGF and NRG stimulationon HER3 and EGFR when they are coexpressed. BecausePAFPs that can be paired with mEos for two-color STORM havesubstantially lower brightness, we focused on single-color ex-periments in which one receptor was tagged with mEos andanother with the monomeric enhanced version of GFP. GFPlabeling allowed the selection of cells that coexpress both re-ceptors but did not interfere with the detection of mEos. We firstlooked at mEos–HER3 coexpressed with EGFR–GFP. Re-markably, HER3 organization was significantly different whencells were stimulated with NRG rather than EGF. In the pres-ence of its own ligand, NRG, and EGFR coexpression, the extentof HER3 clustering was similar to that observed for HER3 aloneupon NRG stimulation (Fig. 2 G and H). We reasoned that thisbehavior is compatible with NRG-induced HER3/EGFR heter-odimers, which in our assay would not be detected as a change inthe HER3 oligomerization state. In contrast, EGF induced sig-nificant clustering of HER3 at the membrane upon EGFRcoexpression (Fig. 2 G and H), but not in the absence of EGFR(Fig. S4A). This effect could be seen as early as 5 min after EGFstimulation and was reminiscent of the behavior of EGFR inresponse to EGF when EGFR was expressed alone (Fig. 2D). In

van Lengerich et al. PNAS | Published online March 20, 2017 | E2839

BIOCH

EMISTR

YPN

ASPL

US

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020

the complementary set of experiments, when mEos–EGFR wascoexpressed with HER3–GFP, we observed that coexpression ofHER3 did not affect EGFR behavior in a measurable way.EGFR still clustered in response to EGF but not in response toNRG (Fig. S4B). EGF-induced HER3 clusters were smaller inreceptor number than EGFR clusters, averaging 5– 12 moleculesin a cluster (Fig. S3B), as determined using the estimate describedabove for EGF-induced EGFR clusters.These data show that although both EGF and NRG are ca-

pable of inducing HER3 phosphorylation in an EGFR-dependentmanner, the underlying organization of HER3 and EGFR com-plexes at the membrane differs quite markedly in response to eachligand. EGFR and HER3 form smaller oligomers, likely dimers, inthe presence of NRG and form higher-order clusters in thepresence of EGF. This difference coincides with the pattern ofEGFR phosphorylation. As with clustering, we observe EGFRphosphorylation only in the presence of EGF, not with NRG,suggesting that clustering of EGFR might be necessary for itsefficient phosphorylation. Because we observed HER3 clusteringonly when HER3 was coexpressed with EGFR and stimulatedwith EGF, we hypothesized that HER3 clustering might be a di-rect consequence of EGF-induced EGFR clustering.

EGFR Clustering Is Necessary for EGF-Dependent HER3 Clustering. Totest our hypothesis that HER3 clustering is a consequence ofEGFR clustering, we first looked at the functional requirementswithin EGFR that contribute to this behavior. We predicted thatan intact kinase is essential for EGFR clustering because EGFRactivation was previously linked to the formation of ligand-induced higher-order oligomers (25, 29, 32, 38). We introduced amutation (V924R) that falls within the asymmetric dimer in-terface between the kinase domains and renders EGFR in-capable of activation (3). We imaged cells expressing the mEos–EGFR–V924R mutant alone by STORM. The V924R mutationcompletely abrogated EGFR clustering in response to ligandbinding across all EGF stimulation time points (Fig. 3 A and D),suggesting that the catalytic activity of EGFR is essential for itsclustering. To ascertain that the inhibitory effect of the V924Rmutation on EGFR clustering is caused by the inhibition ofcatalytic activity and not by the disruption of receptor interac-tions that rely on the interface centered around V924, we

inhibited EGFR activity through an alternative strategy using thesmall-molecule kinase inhibitor gefitinib. Gefitinib binding iscompatible with kinase asymmetric dimerization but blocks ca-talysis (39). The addition of EGF to cells in the presence ofgefitinib also prevented the formation of higher-order EGFRclusters but, interestingly, increased the basal level of clustering(Fig. 3 B and E). We reasoned that this increased basal levelmight be caused by the stabilization of a preformed lower-orderoligomeric state of EGFR, such as a dimer, as has been de-scribed previously for other type I EGFR kinase inhibitors (40).Cumulatively, our results show that EGFR clustering depends onits ability to form a catalytically active complex.We then used the EGFR–V924R mutant to test if HER3

clustering in response to EGF depends on EGFR clustering.When mEos–HER3 was coexpressed with the EGFR–V924R–GFP, we could no longer detect HER3 clustering in response toEGF (Fig. 3 C and F). These results indicate that HER3 re-organization at the plasma membrane in response to EGF islinked directly to ligand-induced changes in the behavior of itsinteraction partner, EGFR.

HER3 Phosphorylation by EGFR Depends on the Ability of EGFR toHomo-Oligomerize. We then examined whether blocking EGFRhomo-oligomerization while retaining its ability to interact withHER3 will influence the extent of HER3 phosphorylation inducedby EGF. To this end, we used the EGFR–V924R mutant. Al-though this mutation blocks the ability of EGFR to self-activateby inhibiting its allosteric activator function, it preserves itsfunction as a receiver kinase. The catalytic function of theEGFR–V924R mutant therefore can be activated by partneringwith a receptor with an intact allosteric activator function, such asHER3 (3, 6).In cells coexpressing EGFR–V924R with HER3, EGFR

phosphorylation could no longer be detected in response to EGF(Fig. 4A). These data indicated that the EGFR phosphorylationdetected upon EGF stimulation is primarily a result of EGFRactivation through self-association and that HER3 cannot actas an allosteric activator receptor for the EGFR kinase underthese circumstances. Moreover, we observed that in these cells,HER3 is no longer phosphorylated in response to EGF (Fig. 4A).This lack of phosphorylation does not result from the inability of

400

300

200

100

0

inte

grat

ed c

ount

s

no ligand NRG EGF

******

200

150

100

50

0

EG

FR

cou

nts

100806040200

distance (nm)

200

150

100

50

0

EG

FR

cou

nts

100806040200

distance (nm)

200

150

100

50

0

HE

R3

coun

ts

100806040200

distance (nm)

mEosEGFR+gefitinib EGFR-V924R-GFP

mEosHER3

EGF

NRGor

mEosEGFR-V924R

EGF EGF

x

mEosmEos mEos

400

300

200

100

0

inte

grat

ed E

GF

R c

ount

s

none 1 min 5 min 10 min

***

***400

300

200

100

0

inte

grat

ed E

GF

R c

ount

s

none 1 min 5 min 10 min

***

*****

GFP

HER3/EGFR

HER3/EGFRV924R

EGFREGFR+gef

EGFREGFR-V924R

10 min EGF5 min EGF1 min EGFno EGF

10 min EGF5 min EGF1 min EGFno EGF

10 min EGF10 min NRGno EGF

A B C

E FD

Fig. 3. Clustering of EGFR and HER3 upon EGF isdependent on EGFR kinase activity. Pairwise distancehistograms calculated as in Fig. 2 for the followingconditions: (A) Cells expressing the mutant mEos–EGFR–V924R were stimulated with 10 nM EGF for theindicated time and were imaged and analyzed bySTORM. (B) Cells expressing mEos–EGFR were treatedwith gefitinib (10 μM) during the entire period ofserum starvation (6 h) and then were stimulated with10 nM EGF. (C) Cells expressing both mEos–HER3 andEGFR–V924R–GFP were stimulated with 10 nM EGF or10 nM NRG for 10 min and then were imaged andanalyzed by STORM. (D–F) Integrated histogramswere calculated from the pair-correlation data asdescribed in Fig. 2 for the mEos–EGFR–V924R mutantand mEos–EGFR with gefitinib. Comparisons withcompiled wild-type mEos–EGFR are shown in D and E,and the comparison with wild-type EGFR–GFP coex-pressed with mEos–HER3 is shown in F. **P < 0.01;***P < 0.001.

E2840 | www.pnas.org/cgi/doi/10.1073/pnas.1617994114 van Lengerich et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020

the EGFR–V924R mutant to phosphorylate HER3, because,upon stimulation with NRG, HER3 was phosphorylated efficientlyby EGFR–V924R (Fig. 4A). Our data therefore suggest that, uponstimulation with EGF, HER3 is phosphorylated by EGFR in amanner that is dependent on the ability of EGFR to form self-activating oligomers.

In the Presence of EGFR and HER3, EGF Preferentially Drives EGFRHomo-Oligomerization. Our data indicate that, in response toNRG, HER3 phosphorylation proceeds via the formation of acanonical EGFR/HER3 heterodimer in which HER3 allosteri-cally activates the EGFR kinase, but such heterodimers mightnot form in the presence of EGF. Rather, in the presence ofEGF, liganded EGFR is much more likely to interact with an-other EGFR. Hence, if both EGF and NRG are present, EGFR/HER3 heterodimers will form only under nonsaturating levels ofEGF when EGF-free EGFR molecules are available to interactwith HER3 in the NRG-stabilized EGFR/HER3 heterodimers(Fig. 4B). To test this idea, we stimulated cells coexpressingHER3 and EGFR–V924R with a saturating concentration ofNRG (10 nM) and a range of EGF concentrations. We predictedthat at low concentrations of EGF an available pool of ligand-free EGFR molecules would be available to form NRG-inducedHER3/EGFR–V924R heterodimers and support HER3 phos-phorylation. Increasing concentrations of EGF should result insequestering EGFR–V924R in dimers, which are catalyticallyinactive because of the inability of EGFR–V924R to form anasymmetric dimer, and therefore HER3 phosphorylation couldno longer be supported. Indeed, in the absence or at low con-centrations of EGF, NRG induced efficient HER3 phosphory-lation by EGFR–V924R, reflecting these two receptors’ ability toheterodimerize efficiently through the asymmetric kinase dimerinterface. In agreement with our prediction, HER3 phosphory-lation decreased progressively with increasing EGF concentra-

tion, indicating a shift in the distribution of NRG-inducedEGFR–V924R/HER3 heterodimers to inactive EGFR–V924Rhomo-oligomers (Fig. 4B).

HER3 Phosphorylation in Response to EGF Proceeds Through aNoncanonical Mechanism. If EGF preferentially drives self-association of EGFR through asymmetric kinase dimerization(Fig. 4B), it would suggest that the chances of forming EGFR/HER3 asymmetric kinase dimers in response to EGF are mini-mal. Nevertheless, we observe that HER3 is robustly phosphor-ylated by the wild-type EGFR in response to EGF (Fig. 1),indicating that its phosphorylation might proceed through amechanism that does not rely on the canonical asymmetric dimerinterface formed between EGFR and HER3. We introduced aV926R mutation in HER3 that is equivalent to the V924Rmutation in EGFR and that disrupts the allosteric activator in-terface between HER3 and EGFR (6). We observed that, in thepresence of EGF, the HER3–V926R mutant is still robustlyphosphorylated on a wide spectrum of phosphorylation sites(Fig. 5A). In the presence of NRG, however, introduction of theV926R mutation in HER3 completely blocks phosphorylation byEGFR (Fig. 5A). These findings demonstrate that EGF-inducedcrosstalk between EGFR and HER3 proceeds through a non-canonical mechanism whereby EGFR and HER3 do not rely onthe formation of a direct asymmetric dimer. NRG induces in-trinsically different interactions between HER3 and EGFR andengages the receptors through asymmetric heterodimerization ofthe kinase domains that relies on HER3’s function as anallosteric activator.Although phosphorylation of HER3 in response to EGF does

not proceed through the canonical dimerization mechanism, it isstrictly dependent on EGFR’s ability to self-activate in EGF-induced homo-oligomers that we observed by STORM (Figs. 2D–F and 4). Our imaging analysis of HER3 coexpressed withEGFR indicates that EGF also mobilizes HER3 to cluster, abehavior that is not observed with the addition of NRG (Fig. 2G–I). We therefore assessed whether the HER3–V926R mutantis recruited to clusters; such recruitment could explain why it stillretains the ability to be phosphorylated by EGFR in response toEGF, reflecting the behavior of wild-type HER3. In agreementwith this assumption, we observed that HER3–V926R is indeedrecruited to clusters efficiently (Fig. 5 B and C), mirroring thebehavior of wild-type HER3. The ability of HER3–V926R tocluster also demonstrates that, like its phosphorylation, HER3clustering in response to EGF is independent of the asymmetricdimerization interface localized in its kinase domain.We subsequently tested if another conserved portion of the

intracellular region of HER receptors, the juxtamembrane seg-ment, could be responsible for mediating the interaction betweenHER3 and EGFR in the EGF-induced clusters and enableHER3 receptor tail phosphorylation. In active HER receptorasymmetric dimers, the juxtamembrane segment of the receiverkinase interacts with a binding site on the kinase domain of theactivator kinase (41). In this asymmetric arrangement, the re-ceiver kinase potentially could bind the juxtamembrane segmentof a third receptor, enabling lateral signaling (Fig. S5). We de-leted the juxtamembrane segment in HER3 to see if this deletionwould disrupt its phosphorylation by EGFR in response to EGF.However, HER3 missing the juxtamembrane segment was stillphosphorylated efficiently by EGFR upon stimulation with EGF(Fig. S5). This finding suggests that HER3 phosphorylation inthe EGF-induced clusters follows a different mechanism thatcould involve a different region of the receptor or might be aresult of a high local concentration of the enzyme (EGFR) andsubstrate (HER3) in the clusters.

HER2 Clusters Together with HER3 in Response to NRG, but TheirSignaling Relies on Asymmetric Kinase Dimerization. In light of thefundamental differences with which ligands control EGFR andHER3 receptor organization, we examined how HER3 interactswith another dimerization partner, HER2. Like HER3, HER2 is

A B

pY1289 HER3

HER3

EGFR

pY1068 EGFR

EGF [nM]:NRG [nM]:

00

0

10

EGFRV924R *

** **

** * *

HER3

EGFNRG

high EGF: EGFR V924Rsequestered in inactive dimers

low EGF: free EGFR V924Ravailable to dimerize with HER3

0.1 1 10 25

EGFR-V924R-GFP

mEos-HER3

EGF

NRGor

x

*

mEos

-EGF NRG

2 5 10 60 2 5 10 60

mEos-HER3:EGFR V924R-GFP: + + + + + + + + +

+ + + + + + + + +

[min]:

pY1289 HER3

HER3

EGFR

pY1068 EGFR

pY1197 HER3

pY1222 HER3

pY1045 EGFR

pY1173 EGFR

mEos-HER3:EGFR V924R-GFP: + + + + + +

+ + + + + +

GFP

Fig. 4. Homo-oligomerization of EGFR is required for HER3 phosphoryla-tion in response to EGF. (A, Upper) Cartoon depicting EGFR–V924R–GFP andmEos–HER3 receptors. (Lower) Western blot analysis of the lysates fromNR6 cells stably expressing mEos–HER3 and EGFR–V924R–GFP. The lysateswere collected after stimulation with 10 nM EGF or 10 nM NRG for the in-dicated periods of time. (B, Upper) The cartoons provide a schematic rep-resentation of predicted interaction preferences between EGFR–V924R andHER3 under different ratios of EGF and NRG. (Lower) Western blot analysisof the lysates from NR6 cells stably expressing mEos–HER3 and EGFR–V924R–GFP. The lysates were collected after stimulation with 10 nM NRG and increasingconcentrations of EGF (0–25 nM) for 10 min. Phosphorylation was monitored atY1289 for HER3 and at Y1068 for EGFR with site-specific antibodies.

van Lengerich et al. PNAS | Published online March 20, 2017 | E2841

BIOCH

EMISTR

YPN

ASPL

US

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020

an obligate heterodimerization partner, because it lacks its ownligands. Therefore, HER2 and HER3 can interact only in re-sponse to HER3-specific ligands. In such complexes, HER2 isassumed to act as the receiver kinase and to phosphorylate HER3.We applied STORM to analyze the behavior of mEos–

HER3 under conditions of HER2–GFP coexpression in theNR6 cells and stimulation with NRG. In contrast to our results inthe analysis of the behavior of the HER3/EGFR complex uponNRG stimulation, membrane organization of HER3 changedsignificantly in response to NRG when HER2 was coexpressed(Fig. 6A). In the presence of HER2, HER3 displayed increasinglevels of clustering upon NRG stimulation, to an extent similarto the HER3 clustering observed in the presence of EGF andEGFR coexpression (Figs. 2 H and I and 6 A and D). To test ifHER3 clustering is accompanied by changes in membrane or-ganization of HER2, we measured the behavior of mEos–HER2 in the presence of HER3 expression. mEos–HER2 alsoclustered significantly in response to NRG, suggesting thatHER2 and HER3 interact upon ligand binding, forming higher-order oligomers at the plasma membrane (Fig. 6 B and E).To investigate the nature of HER3 clusters associated with

HER2 coexpression, we then probed whether they form in-dependently of the asymmetric kinase dimerization interface, aswe observed with EGF-induced HER3 clusters in cells coex-pressing EGFR (Fig. 5B). Introduction of the V926R mutationin HER3 entirely abrogated its clustering upon NRG stimulationwhen HER2 was coexpressed (Fig. 6 C and F). This result sug-gests that these clusters are intrinsically different from theclusters that HER3 forms in the presence of EGFR and EGF.Consequently, breaking the allosteric activator interface by theV926R mutation blocked NRG-dependent HER3 phosphoryla-tion by HER2 (Fig. 6G). Although in our assays we observedNRG-independent HER2 phosphorylation for all HER2 stablecell lines we have generated, HER2 did not engage withHER3 in active complexes in the absence of NRG, as evidencedby the lack of detectable HER3 phosphorylation under theseconditions (Fig. 6G). Taken together, these results show that

HER2/HER3 signaling proceeds through another unique routethat involves higher-order interactions at the plasma membranebut is dependent on the formation of the canonical asymmetrickinase dimer interface.

DiscussionOligomerization of receptor tyrosine kinases, traditionally con-sidered as dimerization, constitutes a basic regulatory principlebehind their activation by extracellular ligands (42). Our obser-vations solidify the notion that EGFR signaling is controlledthrough the formation of oligomeric complexes that extend be-yond a dimer. In agreement with previous studies, we observedthe formation of receptor clusters that increase in size over timeupon EGF stimulation. Our data show that the clusters reach theapproximate size of ∼5–20 receptors upon 10 min of EGFstimulation, corroborating a number associated with spontane-ously formed activated EGFR clusters on the cell surface of thehuman A431 carcinoma cells and perhaps suggesting a self-organizing principle within these clusters (28). We also observethat EGFR clustering is dependent on its kinase activity, aspreviously reported (28, 29), and we show that self-activationthrough the formation of an asymmetric kinase dimer is key toEGFR oligomerization. However, our data indicate that not allactivating stimuli have the same effect on EGFR organization atthe membrane, because when EGFR is recruited to a complexwith another HER receptor, HER3, and activated via aHER3 cognate ligand, it does not cluster.In contrast to EGFR, studies on the regulation of HER3

oligomerization in response to ligands have thus far yielded con-flicting results. Cell-based approaches using indirect measurements,such as chemical crosslinking, nucleic acid aptamers, and bulkfluorescence complementation, provided evidence that HER3 oli-gomerizes before NRG binding, but upon binding the ligand itdissociates into monomeric receptors that are poised to hetero-dimerize with other HER receptors (43–45). More direct ap-proaches, such as immunoelectron microscopy and quantum dotsingle-particle tracking analyses, have challenged this model andpoint to significant HER3 homodimerization and even clustering inresponse to NRG (30, 46). In contrast, no HER3 oligomers havebeen observed during biochemical studies of the isolated extracel-lular domain of HER3 (47, 48) or through studies of chimeric full-length receptors (49).Our findings underscore the importance of considering

HER3 behavior in the context of its molecular environment,which we were able to control by using a cell line that does notendogenously express HER receptors. We observed that, whenexpressed alone, HER3 does not form extensive homo-oligomersthat are dissociated or formed upon ligand stimulation. How-ever, when HER3 is coexpressed with EGFR or HER2, it en-gages in interactions that result in its clustering at the plasmamembrane upon ligand binding. Intriguingly, this behavior isgreatly influenced by the combination of the receptor type andthe type of stimulating ligand. When coexpressed with EGFR,HER3 undergoes significant clustering only in the presence ofEGF but not in the presence of NRG. When HER3 is coex-pressed with HER2, NRG induces significant clustering ofHER3, corroborating previous studies in which measurement ofNRG-dependent HER3 clustering was conducted in cells withmeasurable levels of HER2 expression (30, 46).Most importantly, our results have important implications for

understanding the molecular mechanisms that underlie thecross-communication between HER3 and other HER receptors.In a seemingly reciprocal mode of interaction, such as activationof the HER3/EGFR complex by EGF or NRG, this receptorpair leads to divergent patterns of receptor phosphorylationdepending on the ligand (Fig. 1). Our results show that thesedifferent patterns reflect changes in receptor organization in-duced by different ligands. We show that the lack of EGFRphosphorylation in response to NRG does not reflect its lack ofengagement with HER3, because HER3 is robustly phosphory-lated in an EGFR-dependent manner. Rather, EGFR cannot be

A200

150

100

50

0

HE

R3

coun

ts

100806040200

distance (nm)

no ligand 10 min NRG 10 min EGF

250

200

150

100

inte

grat

ed c

ount

s

no ligand NRG EGF

HER3-VR+EGFR

C

EGFR-GFP

mEos-HER3 V926R

EGF

NRGor

x

mEos

B

*

*

-EGF NRG

2 5 10 60 2 5 10 60

mEos-HER3 V926R:EGFR-GFP: + + + + + + + + +

+ + + + + + + + +

[min]:

pY1289 HER3

HER3

EGFR

pY1068 EGFR

pY1197 HER3

pY1222 HER3

pY1045 EGFR

pY1173 EGFR

GFP

Fig. 5. In the EGF-induced clusters, HER3 is phosphorylated by EGFR inde-pendently of asymmetric kinase dimerization. (A, Upper) Cartoon depictingthe EGFR–GFP and mEos–HER3–V926R receptors. (Lower) Western blotanalysis of the lysates from NR6 cells stably expressing mEos–HER3–V926Rand EGFR–GFP. The lysates were collected after stimulation with 10 nM EGFor 10 nM NRG for the indicated periods of time. (B and C) Pairwise distancehistograms (B) and integrated counts from those histograms calculated fromthe STORM images (C) (as described in Fig. 2) for cells coexpressing themutant mEos–HER3–V926R and EGFR–GFP and stimulated with 10 nM EGF or10 nM NRG for 10 min. *P < 0.05.

E2842 | www.pnas.org/cgi/doi/10.1073/pnas.1617994114 van Lengerich et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020

phosphorylated because NRG is unable to induce homomericinteractions in EGFR. It is not completely clear why EGFRneeds to homo-oligomerize to become phosphorylated efficiently.A preference for the phosphorylation of receptor tails in trans inthe EGFR receptor dimer was reported previously (36) and couldexplain why EGFR phosphorylation is not favorable in an EGFR/HER3 heterodimer in which EGFR would need to phosphorylateits own tail (in cis). Another possibility is that EGFR autophos-phorylation is simply not a very efficient process, and higher-orderclustering of this receptor induced by its own ligands, such as EGF,increases the rate of autophosphorylation. In an analogous man-ner, NRG-dependent clustering of HER2 that we observe is likelynecessary for its efficient phosphorylation as suggested before bythe studies using aptamers that are predicted to selectively blockthese higher-order interactions (50).Asymmetric kinase domain dimerization, with HER3 re-

stricted to the allosteric activator position because of impairedcatalytic activity, has been a benchmark for understanding howHER receptors activate in heteromeric complexes (3, 6). Herewe uncover significant differences in how HER3 forms signalingcomplexes with EGFR and HER2 in response to different li-gands (Fig. 7). In the presence of its own ligand, NRG, HER3engages with its dimerization partners EGFR and HER2 bytaking the function of an allosteric activator kinase, adhering tothe asymmetric kinase dimerization mechanism. However, whenEGFR homomeric complexes form preferentially in the presenceof EGF, HER3 does not form asymmetric kinase dimers withEGFR. Although this phenomenon has been noted before (51),the underlying mechanism was unknown. We show that, uponEGF stimulation, HER3 follows the behavior of its interaction

partner, EGFR, and forms clusters. We propose that this behaviorfacilitates HER3 phosphorylation by promoting activating inter-actions between EGFR and HER3 in which they engage in thekinase/substrate mode rather than kinase/activator mode. Al-though at present we do not know the molecular mechanism ofHER3 clustering in response to EGF-induced EGFR activation,EGFR clustering was shown to be concurrent with the generationand notable rearrangement of anionic lipids in the plasma mem-brane (26, 29). EGFR clustering also was shown to be promotedby depletion of cholesterol, suggesting that the membrane envi-ronment in which EGFR clusters is unlikely to represent lipid rafts(24). These data indicate that biophysical changes in the mem-brane may create an environment conducive to the clustering ofEGFR interaction partners with which EGFR otherwise wouldform only weak interactions. In contrast to EGF, NRG fails toinduce clustering of HER3 and EGFR, but, intriguingly, it doespromote clustering between HER2 and HER3. The importance ofthe asymmetric kinase dimerization in this process suggests thatthis clustering is intrinsically distinct from the clustering ofHER3 and EGFR that we observe upon EGF treatment.In summary, our work brings mechanistic insights into the

previously documented differences in signaling outcomes in-duced by stimulation with EGF and NRG. Our results show that,despite highly conserved structural features of HER receptors,not every ligand induces a receptor heterocomplex that is con-sistent with the asymmetric kinase dimerization mechanism. Inthese cases, higher-order oligomerization and cross-activationthrough lateral interactions in a kinase/substrate-like mode seemto come into play. As a result, there is a qualitative difference inthe phosphorylation states of the engaged receptors, providing

400

300

200

100

0

inte

grat

ed c

ount

s

no ligand 10 min NRG

HER3 HER3/HER2

******

ns

400

300

200

100

0

inte

grat

ed c

ount

s

alone HER3 HER3NRG

HER2***

***

ns

200

150

100

50

0

HE

R3

coun

ts

100806040200

distance (nm)

HER3 HER3 + NRG HER3/HER2 HER3/HER2 + NRG

200

150

100

50

0

HE

R3

coun

ts

100806040200

distance (nm)

HER3/HER2 HER3/HER2 + NRG HER3-VR/HER2 HER3-VR/HER2 + NRG

200

150

100

50

0

HE

R2

coun

ts

100806040200

distance (nm)

HER2 HER2/HER3 HER2/HER3 + NRG

NRG [min]: - 5 10 60 - 2 5 10 60 - 2 5 10 60

HER3

HER2

pY1221/Y1222 HER2

HER2-GFP

mEos-HER3 V926R

x

mEos

*

mEos-HER2

HER3-GFP

x

mEosmEos-HER2

mEosmEos-HER3

HER2-GFP

x

mEos

mEos-HER3

mEos

x

mEos-HER3-V926R:HER2-GFP:

+ + + + + + + + +- + + + + +

mEos-HER3:- - - - - - - -

- - - - -

+ + + + + + + + + +- - - -

pY1289 HER3

GFP GFPGFP

HER3HER3/HER2

HER2

400

300

200

100

0

inte

grat

ed c

ount

s

no ligand 10 min NRG

ns

***

HER3/HER2HER3-VR/HER2

A B C

D

G

E FFig. 6. NRG induces clustering of HER3 in the pres-ence of HER2 and its phosphorylation via the asym-metric kinase dimerization. (A–C, Upper) Cartoonsdepicting the different combinations of mEos–HER3,HER2–GFP, mEos–HER2, and mEos–HER3–V926R usedin the experiments below. (Lower) Pairwise distancehistograms calculated from the STORM images (asdescribed in Fig. 2) for the following conditions.(A) Cells expressing mEos–HER3 alone or mEos–HER3and HER2–GFP and stimulated for 10 min with 10 nMNRG. (B) Cells expressing mEos–HER2 alone or coex-pressing HER3–GFP and stimulated for 10 min with10 nM NRG. (C) Cells coexpressing mEos–HER3–V926Rand HER2–GFP, and stimulated for 10 min with 10 nMNRG. (D–F) Integrated counts of pair-correlation data inA, B, and C, respectively, calculated as described in Fig.2. In F, a relative comparison with the extents of clus-tering to the wild-type mEos–HER3 coexpressed withHER2–GFP is shown. ns, not significant. ***P < 0.001.(G) Western blot analysis of the lysates from NR6 cellsexpressing mEos–HER3 alone, mEos–HER3/HER2–GFP, or mEos–HER3–V926R/HER2–GFP cells were serumstarved 6 h and were stimulated with NRG for 2–60 min.

van Lengerich et al. PNAS | Published online March 20, 2017 | E2843

BIOCH

EMISTR

YPN

ASPL

US

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020

insight into how cells interpret signals delivered by differentHER ligands. It is important to note that relative oligomeriza-tion states of HER receptors in response to ligands are mostlikely influenced by both their stoichiometry and relative ex-pression levels. For example, in the presence of HER3, HER2,and EGFR, NRG might be more likely to favor the formation ofthe HER2/HER3 dimers rather than EGFR/HER3 dimers, be-cause the former have been shown to form preferentially (52).Perturbing the level of expression of any individual receptors andthe relative level of expression among HER family memberscould provide important information about the hierarchy of re-ceptor interactions.Our data also emphasize that HER receptor signaling needs to

be interpreted in the context of higher-order oligomeric struc-tures, in a manner that could be somewhat parallel to ephrin(Eph) receptors. In Eph receptor signaling, first step of ligand-mediated receptor autophosphorylation usually is not sufficientto generate functional responses, and the Eph receptors mustform multimeric complexes to activate downstream signalingrobustly (53). Moreover, the size of the Eph clusters has beencorrelated with the strength and type of the signaling response(53–55). Likewise, our data show that constriction of EGFRreceptor within a heterodimer in which its kinase domain is ef-ficiently activated is still not sufficient for its autophosphor-ylation, and EGFR undergoes robust phosphorylation only whenallowed to cluster. It is tempting to speculate that regulationthrough control of the size of the signaling unit could be alsooperative for other receptor tyrosine kinases to control their netphosphorylation and functional outcomes of their activation.Finally, our findings have important implications for the treat-ment of human diseases in which HER3 signaling plays an im-portant role. HER3 contributes to drug resistance to EGFR- orHER2-targeted therapeutics (56) and recently was discovered tocarry a spectrum of mutations in human cancers (57). The mostrecognized function of this catalytically impaired receptor isserving as an allosteric activator of other HER receptors. Ourdata show that, under some conditions, HER3 can contribute tosignaling independently of this function, emphasizing the needfor the careful design of therapeutic strategies targeting HER3.

Materials and MethodsPlasmids and Cell Lines.Murine stem cell virus (MSCV) plasmids and Plat-e cellswere a generous gift from theWells laboratory at the University of California,San Francisco (UCSF), and NR6 cells were a generous gift from the Moasser

laboratory at UCSF. mEos3.2 was fused between the human EGFR, HER2, orHER3 signal peptide domains, residues 1–24 (EGFR), 1–22 (HER2), 1–19(HER3), and the remainder of the sequence using a homemade Gibson as-sembly reagent into retrovirus-capable vectors (MSCV). Similarly, for mEos-tagged PH constructs, one, two, or three copies of mEos3.2 were fused at theN terminus of the PH domain of phospholipase C delta 1 (PLCΔ); for theCD86 construct, mEos3.2 was fused at the C terminus of the gene. Virus wasproduced by transient transfection of the plasmid into Plat-e cells usingFuGENE (Promega), and virus-containing supernatant was collected after2 d. Supernatant was filtered (0.22-μm pore diameter) and was added im-mediately to NR6 cells plated on the previous day; antibiotic (250 μg/mLhygromycin or 2 μg/mL puromycin) was added 24 h after viral infection. Cellswere cultured in Gibco DMEM supplemented with 10% FBS, penicillin/streptomycin, sodium pyruvate, and nonessential amino acids. For cellsexpressing two different HER receptors, one receptor was tagged withmonomeric EGFP at the C terminus after the terminal residue of the HER re-ceptor (e.g., EGFR–GFP), and the other receptor was tagged with mEos at theN terminus after the signal peptide (e.g., mEos–HER3). The second constructalso was introduced by stable transfection (as above) to a cell line alreadyexpressing the mEos-fused receptor, and coexpressing cells were selected byadding both hygromycin (250 μg/mL) and puromycin (2 μg/mL) to the medium.

Sample Preparation. Eight-well chambered coverglass slides (Lab-Tek) werewashed with 1 M KOH for 10 min and then were washed and coated withpolylysine (0.01%) for 30 min. Stable cell lines expressing the proteins ofinterest were deposited on the washed glass slide and allowed to adhere for36 h. Cells were serum starved for 6 h and then were stimulated for specifiedamounts of time with 10 nM EGF or NRG (saturating concentration) at 37 °C.Cells were fixed with 4% formaldehyde for 10 min at 20 °C, washed, andkept in PBS at 4 °C for up to 1 wk in the dark before imaging.

Microscopy. Fixed cells were imaged using a home-built STORM invertedmicroscope consisting of 405-nm, 488-nm, 561-nm, and 647-nm lasers. Thelasers were coaligned and reflected by a quad-band dichroic mirror through a10× Olympus objective (NA 1.4), and fluorescence was collected from thesame objective with an Andor EMCCD camera. Images were collected andprocessed with a home-written software program. Maximum laser powerused during STORM measured before the objective was 0.03 mW for 405 nmand 65 mW for 561 nm. Cells expressing only one receptor fused to mEoswere carefully located at low green intensity using a wide-field setting(20 Hz) to minimize bleaching of mEos. The cells then were illuminated witha high intensity (65 mW) 561-nm laser at 60 Hz, with an activation frame(405 nm) every 10 frames. By increasing the power on the 405-nm laser (zeroto 0.03 mW) during imaging, mEos was photoconverted from a green to a redstate and could be observed as single fluorophores. The activation frames wereprogrammed to occur simultaneously with a brightfield image, which was usedin postprocessing to correct for drift. Approximately 10,000–60,000 imageswere collected, until all mEos fluorophores were photobleached, while the405-nm power was kept low enough to activate mEos only sparsely (about5–30 fluorophores per frame) to avoid improper counting in the analysis. Forcells expressing two receptors (e.g., mEos–HER3/EGFR–GFP), the GFP was muchbrighter than mEos, and cells could be located easily with low intensity in thegreen channel. Cells then were subjected to the STORM imaging, and if fluo-rophores appeared in the red channel, it was clear that cells coexpressed bothreceptors. In some cases, cells were located that did not contain mEos but didcontain GFP; these cells were disregarded. Generally, cells were imaged only ifthey were within the range of expression that was appropriate for the analysis,20–60 molecules/μm2, which is the value of the blink-corrected pair-correlationfunction at long distances (>1 μm). At lower expression levels the signal tobackground was insufficient, and at higher expression levels the movies becameexceedingly long (more than 20 min), at which point autofluorescence caused bythe 405-nm laser illumination could become problematic. The details of STORMimage analysis is described in SI Materials and Methods.

Signaling Assays. NR6 cells stably expressing HER receptors were plated at adensity of 120,000 cells per 10-cm round plate andwere allowed to adhere for36 h. Cells were serum starved and stimulated with 10 nM (or as specified forthe titration) ligand for specified amounts of time at 37 °C. Cells then wereset on ice and lysed and collected in Tris (50 mM)/NaCl (150 mM) buffercontaining 1% Triton X-100, 1 mM Na3VO4, 1 mM EDTA, 1 mM NaF, andone-fourth of a cOmplete Mini, EDTA-free protease inhibitor tablet (Roche)per 10 mL buffer. The lysates were spun down to remove larger organelles,and the supernatant was assayed for protein concentration using a BCA kit(Thermo Fisher). Samples were run on SDS/PAGE, transferred to a PVDF mem-brane, and blotted for with antibodies (see Table S1 for a list of antibodies).

EGFR phosphorylation proceeds via the asymmetric kinase dimer, but HER3 is phosphorylated via a different mechanism,which involves receptor clustering

HER3 phosphorylation by EGFR and HER2 depends on asymmetric kinase dimerization,and involves receptor clustering only in thepresence of HER2

HER3EGFR

PP

P

x

HER3EGFR HER2

EGF NRG

P

P

P

PP

P

PP

PPP

P

P

PP

P

x

PP

P

x

P

P

HER3

PP

P

PP

P

x

PP

P

x

P

PP

P

x

P

P

Fig. 7. Summary of the underlying mechanistic differences in signaling byHER3-containing heteromeric complexes. (Left) Upon stimulation with EGF,EGFR clusters and self-activates through asymmetric kinase dimerization.Under these conditions, HER3 also clusters and is phosphorylated by EGFRwithout engaging with it through the asymmetric kinase dimer interface.(Right) NRG does not lead to clustering of either EGFR or HER3 when theyare coexpressed and drives EGFR kinase activation through asymmetric ki-nase heterodimerization with HER3. In these dimers, HER3 plays a role of theallosteric activator. When HER2 and HER3 are coexpressed, they both clusterin response to NRG, and their signaling is dependent on HER3’s function asan allosteric activator of HER2.

E2844 | www.pnas.org/cgi/doi/10.1073/pnas.1617994114 van Lengerich et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020

After secondary incubation with HRP-linked antibody, ECL prime Westernblotting detection agent (GE Healthcare) was added to the membrane, andmembranes were imaged on a developer.

ACKNOWLEDGMENTS. We thank Z. Gartner, J. Kung, N. Michael, andT. M. Thaker for critical reading of the manuscript and insightful discussions,

S. Liang for many valuable insights; and R. McGorty and V. Pessino from theHuang laboratory for valuable assistance with STORM imaging and helpfuldiscussions. This work was supported by American Cancer Society Postdoc-toral Fellowship 124801-PF-13-365-01-TBE (to B.v.L.), National Institute ofGeneral Medical Sciences Grant R01 GM109176 (to N.J.), and National Insti-tutes of Health New Innovator Award (DP2 OD008479 (to B.H.).

1. Yarden Y, Sliwkowski MX (2001) Untangling the ErbB signalling network. Nat Rev MolCell Biol 2(2):127–137.

2. LemmonMA, Schlessinger J, Ferguson KM (2014) The EGFR family: Not so prototypicalreceptor tyrosine kinases. Cold Spring Harb Perspect Biol 6(4):a020768.

3. Zhang X, Gureasko J, Shen K, Cole PA, Kuriyan J (2006) An allosteric mechanism foractivation of the kinase domain of epidermal growth factor receptor. Cell 125(6):1137–1149.

4. Qiu C, et al. (2008) Mechanism of activation and inhibition of the HER4/ErbB4 kinase.Structure 16(3):460–467.

5. Littlefield P, et al. (2014) Structural analysis of the EGFR/HER3 heterodimer reveals themolecular basis for activating HER3 mutations. Sci Signal 7(354):ra114.

6. Jura N, Shan Y, Cao X, Shaw DE, Kuriyan J (2009) Structural analysis of the catalyticallyinactive kinase domain of the human EGF receptor 3. Proc Natl Acad Sci USA 106(51):21608–21613.

7. Monsey J, Shen W, Schlesinger P, Bose R (2010) Her4 and Her2/neu tyrosine kinasedomains dimerize and activate in a reconstituted in vitro system. J Biol Chem 285(10):7035–7044.

8. Olayioye MA, et al. (1998) ErbB-1 and ErbB-2 acquire distinct signaling propertiesdependent upon their dimerization partner. Mol Cell Biol 18(9):5042–5051.

9. Riese DJ, 2nd, van Raaij TM, Plowman GD, Andrews GC, Stern DF (1995) The cellularresponse to neuregulins is governed by complex interactions of the erbB receptorfamily. Mol Cell Biol 15(10):5770–5776.

10. Pinkas-Kramarski R, et al. (1996) Diversification of Neu differentiation factor andepidermal growth factor signaling by combinatorial receptor interactions. EMBO J15(10):2452–2467.

11. Ueno Y, et al. (2008) Heregulin-induced activation of ErbB3 by EGFR tyrosine kinaseactivity promotes tumor growth and metastasis in melanoma cells. Int J Cancer 123(2):340–347.

12. Carver RS, Stevenson MC, Scheving LA, Russell WE (2002) Diverse expression of ErbBreceptor proteins during rat liver development and regeneration. Gastroenterology123(6):2017–2027.

13. Choi SH, Mendrola JM, Lemmon MA (2007) EGF-independent activation of cell-surface EGF receptors harboring mutations found in gefitinib-sensitive lung cancer.Oncogene 26(11):1567–1576.

14. Crovello CS, Lai C, Cantley LC, Carraway KL, 3rd (1998) Differential signaling by theepidermal growth factor-like growth factors neuregulin-1 and neuregulin-2. J BiolChem 273(41):26954–26961.

15. Yarden Y, Schlessinger J (1987) Epidermal growth factor induces rapid, reversibleaggregation of the purified epidermal growth factor receptor. Biochemistry 26(5):1443–1451.

16. Yarden Y, Schlessinger J (1987) Self-phosphorylation of epidermal growth factor re-ceptor: Evidence for a model of intermolecular allosteric activation. Biochemistry26(5):1434–1442.

17. Sako Y, Minoghchi S, Yanagida T (2000) Single-molecule imaging of EGFR signallingon the surface of living cells. Nat Cell Biol 2(3):168–172.

18. Whitson KB, Beechem JM, Beth AH, Staros JV (2004) Preparation and characterizationof Alexa Fluor 594-labeled epidermal growth factor for fluorescence resonance en-ergy transfer studies: Application to the epidermal growth factor receptor. AnalBiochem 324(2):227–236.

19. Gadella TW, Jr, Jovin TM (1995) Oligomerization of epidermal growth factor recep-tors on A431 cells studied by time-resolved fluorescence imaging microscopy. A ste-reochemical model for tyrosine kinase receptor activation. J Cell Biol 129(6):1543–1558.

20. Ichinose J, Murata M, Yanagida T, Sako Y (2004) EGF signalling amplification inducedby dynamic clustering of EGFR. Biochem Biophys Res Commun 324(3):1143–1149.

21. Needham SR, et al. (2016) EGFR oligomerization organizes kinase-active dimers intocompetent signalling platforms. Nat Commun 7:13307.

22. Clayton AH, et al. (2005) Ligand-induced dimer-tetramer transition during the acti-vation of the cell surface epidermal growth factor receptor-A multidimensional mi-croscopy analysis. J Biol Chem 280(34):30392–30399.

23. Kozer N, et al. (2013) Exploring higher-order EGFR oligomerisation and phosphory-lation–a combined experimental and theoretical approach. Mol Biosyst 9(7):1849–1863.

24. Saffarian S, Li Y, Elson EL, Pike LJ (2007) Oligomerization of the EGF receptor in-vestigated by live cell fluorescence intensity distribution analysis. Biophys J 93(3):1021–1031.

25. Huang Y, et al. (2016) Molecular basis for multimerization in the activation of theepidermal growth factor receptor. eLife 5:5.

26. Wang Y, et al. (2014) Regulation of EGFR nanocluster formation by ionic protein-lipidinteraction. Cell Res 24(8):959–976.

27. Clayton AH, Orchard SG, Nice EC, Posner RG, Burgess AW (2008) Predominance ofactivated EGFR higher-order oligomers on the cell surface. Growth Factors 26(6):316–324.

28. Clayton AH, Tavarnesi ML, Johns TG (2007) Unligated epidermal growth factor re-ceptor forms higher order oligomers within microclusters on A431 cells that aresensitive to tyrosine kinase inhibitor binding. Biochemistry 46(15):4589–4597.

29. Ariotti N, et al. (2010) Epidermal growth factor receptor activation remodels theplasma membrane lipid environment to induce nanocluster formation. Mol Cell Biol30(15):3795–3804.

30. Yang S, et al. (2007) Mapping ErbB receptors on breast cancer cell membranes during

signal transduction. J Cell Sci 120(Pt 16):2763–2773.31. Nagy P, Claus J, Jovin TM, Arndt-Jovin DJ (2010) Distribution of resting and ligand-

bound ErbB1 and ErbB2 receptor tyrosine kinases in living cells using number and

brightness analysis. Proc Natl Acad Sci USA 107(38):16524–16529.32. Chung I, et al. (2010) Spatial control of EGF receptor activation by reversible di-

merization on living cells. Nature 464(7289):783–787.33. Low-Nam ST, et al. (2011) ErbB1 dimerization is promoted by domain co-confinement

and stabilized by ligand binding. Nat Struct Mol Biol 18(11):1244–1249.34. Puchner EM, Walter JM, Kasper R, Huang B, Lim WA (2013) Counting molecules in

single organelles with superresolution microscopy allows tracking of the endosomematuration trajectory. Proc Natl Acad Sci USA 110(40):16015–16020.

35. Kawamoto T, et al. (1991) Quantitative assay of epidermal growth factor receptor inhuman squamous cell carcinomas of the oral region by an avidin-biotin method. Jpn JCancer Res 82(4):403–410.

36. Kovacs E, et al. (2015) Analysis of the role of the C-terminal tail in the regulation ofthe epidermal growth factor receptor. Mol Cell Biol 35(17):3083–3102.

37. Fricke F, Beaudouin J, Eils R, Heilemann M (2015) One, two or three? Probing thestoichiometry of membrane proteins by single-molecule localization microscopy. Sci

Rep 5:14072.38. Abulrob A, et al. (2010) Nanoscale imaging of epidermal growth factor receptor

clustering: Effects of inhibitors. J Biol Chem 285(5):3145–3156.39. Yun CH, et al. (2007) Structures of lung cancer-derived EGFR mutants and inhibitor

complexes: Mechanism of activation and insights into differential inhibitor sensitivity.Cancer Cell 11(3):217–227.

40. Gan HK, et al. (2007) The epidermal growth factor receptor (EGFR) tyrosine kinaseinhibitor AG1478 increases the formation of inactive untethered EGFR dimers. Im-plications for combination therapy with monoclonal antibody 806. J Biol Chem 282(5):2840–2850.

41. Jura N, et al. (2009) Mechanism for activation of the EGF receptor catalytic domain bythe juxtamembrane segment. Cell 137(7):1293–1307.

42. Lemmon MA, Schlessinger J (2010) Cell signaling by receptor tyrosine kinases. Cell141(7):1117–1134.

43. Macdonald-Obermann JL, Adak S, Landgraf R, Piwnica-Worms D, Pike LJ (2013) Dy-

namic analysis of the epidermal growth factor (EGF) receptor-ErbB2-ErbB3 proteinnetwork by luciferase fragment complementation imaging. J Biol Chem 288(42):30773–30784.

44. Park E, Baron R, Landgraf R (2008) Higher-order association states of cellularERBB3 probed with photo-cross-linkable aptamers. Biochemistry 47(46):11992–12005.

45. Kani K, Warren CM, Kaddis CS, Loo JA, Landgraf R (2005) Oligomers of ERBB3 have

two distinct interfaces that differ in their sensitivity to disruption by heregulin. J BiolChem 280(9):8238–8247.

46. Steinkamp MP, et al. (2014) erbB3 is an active tyrosine kinase capable of homo- andheterointeractions. Mol Cell Biol 34(6):965–977.

47. Cho HS, Leahy DJ (2002) Structure of the extracellular region of HER3 reveals an in-terdomain tether. Science 297(5585):1330–1333.

48. Ferguson KM, Darling PJ, MohanMJ, Macatee TL, LemmonMA (2000) Extracellular domainsdrive homo- but not hetero-dimerization of erbB receptors. EMBO J 19(17):4632–4643.

49. Berger MB, Mendrola JM, Lemmon MA (2004) ErbB3/HER3 does not homodimerizeupon neuregulin binding at the cell surface. FEBS Lett 569(1-3):332–336.

50. Zhang Q, Park E, Kani K, Landgraf R (2012) Functional isolation of activated andunilaterally phosphorylated heterodimers of ERBB2 and ERBB3 as scaffolds in ligand-

dependent signaling. Proc Natl Acad Sci USA 109(33):13237–13242.51. Kancha RK, von Bubnoff N, Duyster J (2013) Asymmetric kinase dimer formation is

crucial for the activation of oncogenic EGFRvIII but not for ERBB3 phosphorylation.

Cell Commun Signal 11:39.52. Tzahar E, et al. (1996) A hierarchical network of interreceptor interactions determines

signal transduction by Neu differentiation factor/neuregulin and epidermal growthfactor. Mol Cell Biol 16(10):5276–5287.

53. Davis S, et al. (1994) Ligands for EPH-related receptor tyrosine kinases that requiremembrane attachment or clustering for activity. Science 266(5186):816–819.

54. Egea J, et al. (2005) Regulation of EphA 4 kinase activity is required for a subset ofaxon guidance decisions suggesting a key role for receptor clustering in Eph function.Neuron 47(4):515–528.

55. Salaita K, et al. (2010) Restriction of receptor movement alters cellular response:Physical force sensing by EphA2. Science 327(5971):1380–1385.

56. Sergina NV, et al. (2007) Escape from HER-family tyrosine kinase inhibitor therapy by

the kinase-inactive HER3. Nature 445(7126):437–441.57. Jaiswal BS, et al. (2013) Oncogenic ERBB3 mutations in human cancers. Cancer Cell

23(5):603–617.

van Lengerich et al. PNAS | Published online March 20, 2017 | E2845

BIOCH

EMISTR

YPN

ASPL

US

Dow

nloa

ded

by g

uest

on

Oct

ober

6, 2

020